Class AIReviewData

constructor

• new AIReviewData(socket, move_tree, ai_review, game_id): AIReviewData

• Parameters

  • socket: GobanSocket<ClientToAIServer, AIServerToClient>
  • move_tree: MoveTree
  • ai_review: JGOFAIReview
  • game_id: string | number

  Returns AIReviewData

Readonly move_tree

Readonly socket

Readonly uuid

analyzed_variations

• get analyzed_variations(): undefined | {
      [key: string]: JGOFAIReviewMove;
  }

• Analysis of variations in the game.

  Returns undefined | {
      [key: string]: JGOFAIReviewMove;
  }

date

• get date(): number

• millisecond epoch time (ms from 1970 UTC)

  Returns number

engine

• get engine(): string

• Returns string

engine_version

• get engine_version(): string

• Returns string

error

• get error(): undefined | {
      coords: string;
      message_id: GobanMoveErrorMessageId;
      move_number: number;
  }

• If there was an error processing the review, it can be stored here

  Returns undefined | {
      coords: string;
      message_id: GobanMoveErrorMessageId;
      move_number: number;
  }

id

• get id(): string

• Returns string

moves

• get moves(): {
      [key: string]: JGOFAIReviewMove;
  }

• Analysis of moves in the game.

  Returns {
      [key: string]: JGOFAIReviewMove;
  }

  • [key: string]: JGOFAIReviewMove

network

• get network(): string

• Returns string

network_size

• get network_size(): string

• Returns string

scores

• get scores(): undefined | number[]

• predicted scores that black will win or lose by (negative for loss) for all moves

  Returns undefined | number[]

strength

• get strength(): number

• Returns number

type

• get type(): "fast" | "full"

• A fast review typically only has a few moves reviewed, whereas a full review is expected to have every move reviewed. Note that this sets an expectation but not a requirement on what values are stored in moves, and while games are being reviewed these objects will have zero or more entries in moves regardless of the type.

  Returns "fast" | "full"

win_rate

• get win_rate(): number

• predicted probability that black will win the last move

  Returns number

win_rates

• get win_rates(): number[]

• predicted probability that black will win for all moves

  Returns number[]

analyze_variation

• analyze_variation(uuid, game_id, ai_review_id, cur_move, trunk_move): void

• Parameters

  • uuid: string
  • game_id: number
  • ai_review_id: number
  • cur_move: MoveTree
  • trunk_move: MoveTree

  Returns void

categorize

• categorize(engine, thresholds?, include_negative_score_loss?): null | AiReviewCategorization

• Parameters

  • engine: GobanEngine
  • thresholds: ScoreDiffThresholds = DEFAULT_SCORE_DIFF_THRESHOLDS
  • include_negative_score_loss: boolean = false

  Returns null | AiReviewCategorization

destroy

• destroy(): void

• Returns void